linux-kernel - Re: [PATCH] mm: Drop unused set_pte

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <ZuB_524Dd1tezlX6@casper.infradead.org>
Date: Tue, 10 Sep 2024 18:20:39 +0100
From: Matthew Wilcox <willy@...radead.org>
To: Ryan Roberts <ryan.roberts@....com>
Cc: Anshuman Khandual <anshuman.khandual@....com>, linux-mm@...ck.org,
	Andrew Morton <akpm@...ux-foundation.org>,
	David Hildenbrand <david@...hat.com>,
	"Mike Rapoport (IBM)" <rppt@...nel.org>,
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH] mm: Drop unused set_pte_safe()

On Tue, Sep 10, 2024 at 10:08:10AM +0100, Ryan Roberts wrote:
> On 10/09/2024 10:04, Anshuman Khandual wrote:
> > All set_pte_safe() usage have been dropped after the commit eccd906484d1
> > ("x86/mm: Do not use set_{pud, pmd}_safe() when splitting a large page")
> > This just drops now unused helper set_pte_safe().
> 
> It would be good to add some comment here to mention that the macro was buggy
> due to doing direct dereferencing of the pte, and that if it were to be kept, it
> should have been updated to use a single call to ptep_get().

I'm not sure that the _macro_ would have been buggy in such a scenario.
If I understand correctly, the _caller_ would have been buggy:

/*
 * Use set_p*_safe(), and elide TLB flushing, when confident that *no*
 * TLB flush will be required as a result of the "set". For example, use
 * in scenarios where it is known ahead of time that the routine is
 * setting non-present entries, or re-setting an existing entry to the
 * same value. Otherwise, use the typical "set" helpers and flush the
 * TLB.
 */

so if *ptep changes between the two calls, that's the caller's bug,
right?

Otherwise, set_pmd_safe() would be buggy ...

> With that:
> 
> Reviewed-by: Ryan Roberts <ryan.roberts@....com>
> 
> Thanks,
> Ryan
> 
> > 
> > Cc: Andrew Morton <akpm@...ux-foundation.org>
> > Cc: David Hildenbrand <david@...hat.com>
> > Cc: Ryan Roberts <ryan.roberts@....com>
> > Cc: "Mike Rapoport (IBM)" <rppt@...nel.org>
> > Cc: linux-mm@...ck.org
> > Cc: linux-kernel@...r.kernel.org
> > Signed-off-by: Anshuman Khandual <anshuman.khandual@....com>
> > ---
> >  include/linux/pgtable.h | 6 ------
> >  1 file changed, 6 deletions(-)
> > 
> > diff --git a/include/linux/pgtable.h b/include/linux/pgtable.h
> > index 2a6a3cccfc36..aeabbf0db7c8 100644
> > --- a/include/linux/pgtable.h
> > +++ b/include/linux/pgtable.h
> > @@ -1058,12 +1058,6 @@ static inline int pgd_same(pgd_t pgd_a, pgd_t pgd_b)
> >   * same value. Otherwise, use the typical "set" helpers and flush the
> >   * TLB.
> >   */
> > -#define set_pte_safe(ptep, pte) \
> > -({ \
> > -	WARN_ON_ONCE(pte_present(*ptep) && !pte_same(*ptep, pte)); \
> > -	set_pte(ptep, pte); \
> > -})
> > -
> >  #define set_pmd_safe(pmdp, pmd) \
> >  ({ \
> >  	WARN_ON_ONCE(pmd_present(*pmdp) && !pmd_same(*pmdp, pmd)); \
> 
>